Uncoupling electromagnets

[richbrummitt]

I've made my trackbed really deep (nearly and inch in total) and remembered yesterday that I need to put uncoupling magnets in. I want electromagnets and I have a stash of Seep point motors. What I was thinking was to butcher them and use the long core with just one end coil. The long core could finish just below rail level and allow the coil to sit below the trackbed. I think all I have to do to make this work is permanently fix the core and coil together so they don't fly off. Is my theory sound? Will it work?

[davknigh]

You might find the following link helpful; http://dingham.co.uk/DIY%20magnet.htm. I take it that you are using similar types of couplings to Dinghams such as AJ's or Spratt & Winkle? This type of electromagnet will not work with Kadees.

 

HTH

 

David

[richbrummitt]

You might find the following link helpful; http://dingham.co.uk/DIY%20magnet.htm. I take it that you are using similar types of couplings to Dinghams such as AJ's or Spratt & Winkle? This type of electromagnet will not work with Kadees.

 

HTH

 

David

 

Thanks David, Link didn't work (but neither do some of the others on the Dingham site). I'm using AJ couplings.

[Donw]

If you can find a suitable nail to push the coil onto you could extend the pole quite a bit. John Lagham (think I've got the name right) did a useful bit on electromagnets in the old Model Railway News probably in the 60's and I am sure he would have been using AJs. I've prpbably got it somewhere.

Donw

[Dave at Honley Tank]

If you can find a suitable nail to push the coil onto you could extend the pole quite a bit. John Lagham (think I've got the name right) did a useful bit on electromagnets in the old Model Railway News probably in the 60's and I am sure he would have been using AJs. I've prpbably got it somewhere.

Donw

 

Actually it was John Langan, now unfortunately no longer with us.

John & his friends at Manchester MRS most usually used a then very readably available electrical relay known as a "GPO relay". These are no longer in telephonic use but pr-used ones can still be found for purchase at negligible cost. A 4" round wire nail can be utilised as the soft-iron core of a home made electro-magnet and winding about 5000turns of 36SWG insulated wire as its coil, and energising this at 24Vdc will give some 1200ampere-turns which is adequate as an Alex Jackson uncoupling magnet. The nails cost little from B&Q or similar and Maplins can fix you up with the wire. It does mean a separate transformer-rectifier for your uncouple circuit but it is anyway wise to use a totally separate supply to that which drives your trains.

More complete details in the book "Alex Jackson, the man and the coupling."

[Coombe Barton]

Having just looked at the prices being charged for ex GPO relayis on eBay the Wizard Models ones are less cost. http://www.modelsignals.com/s&w_detail/dg/pk_mag_page.htm

 

They've been confirmed as OK for AJ couplings - although I have not yet tried them. My order is going in shortly.

[Donw]

Thankyou Dave for the info on John Langan. Thinking about winding your own coils we could do with one suitable for 16vac I don't suppose you have the details?

Donw

[richbrummitt]

Having just looked at the prices being charged for ex GPO relayis on eBay the Wizard Models ones are less cost. http://www.modelsign...pk_mag_page.htm

 

They've been confirmed as OK for AJ couplings - although I have not yet tried them. My order is going in shortly.

 

I think I'm going this way too. The gaugemaster uncoupler pin isn't long enough - I've got a ½" track bed with a 3mm ply 'underlay' on top so I need quite a long pin. 14mm should be just about enough I think, and they're cheaper.

[bertiedog]

With AJ couplings I would have thought 5000 turns is a bit high for the coil on a steel nail core, from memory I think mine were about 3000 turns.

 

If 5000 turns are used it makes it more universal. To wind a power screwdriver is useful. also old crt tv's, or monitors, can provide the wire from the tube coils. Another useful core is old aerial ferrite cores, 6mm diameter, the material retrieved from old radios, or bought cheap on Ebay. The turns can be much less with ferrite.

 

Stephen

[Dave at Honley Tank]

Thankyou Dave for the info on John Langan. Thinking about winding your own coils we could do with one suitable for 16vac I don't suppose you have the details?

Donw

 

 

Sorry Don no, not without searching old books for tech.info about insulated wire and then sitting down with a calculator and doing a pile-o-'sums' which would need to include the average diameter and the length of the coil you were going to wind

My advice is certainly to forget using ac for any uncoupler solenoid. A.c. produces an alternating magnetic field, not only alternating but being switched off 100 times a second; your coupling is likely to sit there dithering rather than being firmly attracted to the magnet's core. If you already have the 16vac supply, at least use a bridge rectifier. For this job a 2A rated rectifier will be more than ample and our good friends at Maplins sell those for a very small outlay.

Unless you are really strapped for cash then I think you would be best advised to try the solenoids from Model Signals as 'richbrummitt' suggested. Andrew there is a good bloke and can give further advice.

'bertiedog'thinks he got away with 3000turns but does not quote coil resistance or operating voltage so we can't calculate the ampere-turns of his system.

Whatever the coil, what ever the voltage used, you need to be aiming for an ampere-turn figure of the order of 1200, if AJs are going to work OK. So Stephen, I think your 3000turn coil draws around 0.4A when energised.

Scratch-building this type of solenoid tends to be a bit "suck-it-and-see" and it does need a bit of electrical knowledge and the only firm figure I can be sure of is that 1200 ampere-turns. John Langan got that with the figures I quoted earlier.

[Donw]

Thanks anyway Dave. I normally use three links in 0 but I was thinking of DGs in 2mm.

Donw

[Dave at Honley Tank]

Since my last input on this thread, I’ve discovered an electro-magnet in my ‘bits & pieces’ box. It appears to be one that I must have purchased from Model Signal Engineering at some time, intending to test it for use with Alex Jackson couplings and then totally forgetting that I had got it! It meets the dimensions quoted on the Model Signal Engineering/Wizard Models web page.

 

The unit has a label around the coil which says …â€PK electro-magnet for intermittent use on 6-12Vdcâ€. The coil resistance is about 4 ohms meaning it should drag about 3A from a 12V supply, but short of totally dismantling the coil and counting the turns as I unwound it, I cannot state its ampere-turn value. I suspect that it is a good bit less than 1200 when operated on 12V.

 

I lashed up a simple test rig with the unit mounted with the top of its pole piece just a touch below sleeper height but the track length was so short I could only test on one wagon at a time.

 

My source of supply was a 12V transformer and a 2A bridge rectifier, which gave me a measured 11.55Vdc. With this connected to the coil it did indeed demand about 3A and it nicely pulled down the Alex Jackson coupling on the test vehicle. Left connected for about a minute made the coil very warm, if not hot , to touch, so the “intermittent use†must be followed.

 

However lowering the voltage to 6Vdc caused the magnet to become too weak to attract the coupling. So the likelihood is that the magnet is indeed OK for AJs but more rigorous testing on a working layout over many hours will be necessary before I would be prepared to state categorically that this type of magnet could be used with confidence.

 

Indeed, I later extended the track length on the test rig and used two coupled wagons. The effect of the side thrust of each coupling wire against the other added sufficient extra resistance to magnetic attraction that uncoupling was not always achieved with the ’12’ V supply. A more robust transformer capable of coping with the 3A demand and therefore having less internal voltage drop, and perhaps adding a smoothing circuit could perhaps give more consistent results. Perhaps even firing at 24V would be OK if the “intermittent†was very strictly observed!

 

Please remember that all this was a quick, lash-up of a testing sequence so the ‘jury is still out’

[Dave at Honley Tank]

Since my last input here I’ve searched out two differing transformers, both capable of giving 5A, one at 12V and one at 24V. I also found a 5A rectifier and carried out some more test on my lashed-up test rig using a Wizard Models electro-magnet and two wagons taken straight from the layout; i.e. the couplings were not deliberately set up correctly on my Alex Jackson coupling test jig.

 

The measured full-wave rectified voltage from the 24Vac was 22.5Vdc and the energised magnet took about 5A (the ammeter was 5A FSD and the needle was up against the stop!) – say 5.5A. Every uncoupling attempt was successful. However I also did a simple ‘touch’ temperature test of the coil and it was warm to the touch. After that test session, I allowed the coil to cool (meaning I went back into the dinning room for a cup of tea. )

Next test was to energise the coil at this 22.5Vdc for 10 seconds. The coil was distinctly hot to the touch and it was possible to smell hot insulation.

 

Conclusion? A rectified 24V will certainly give you good Alex Jackson uncoupling but there is a great need to ensure that the coil cannot remain energised and there is still a risk of coil burn-out over an extended shunting session or with a heavy-handed, untrained operator. Not really to be recommended.

 

With the 12V transformer, the rectifier gave me a measured 10.6Vdc without any load, but the voltage measured when the coil was energised had fallen to 8.6Vdc,

and the current was fairly steady at a little less than 3A. The coil was slightly warm to the touch after the uncoupling tests (about 10). All uncoupling attempts were successful.

 

Conclusion? I would be prepared to use this design of electro-magnet as an AJ uncoupler, operated with a 12Vac-rectified supply but with a transformer not less than 50VA continuous rating. Until I had done so over many hours of operation both at home and under exhibition conditions I would however not be prepared to say that I had found the best available answer for AJ uncouplers. For me that is still a suitably selected GPO relay with its contacts removed,or a purpose wound, home made unit giving at least 1200ampere-turns.

 

Actually, as my current new layout (Wheegram Sidings) utilises a track-bed of 2†thick wall insulation, I will probably need to use a home-made magnet with an extended iron core.

[Major Clanger]

 

The unit has a label around the coil which says …â€PK electro-magnet for intermittent use on 6-12Vdcâ€. The coil resistance is about 4 ohms meaning it should drag about 3A from a 12V supply, but short of totally dismantling the coil and counting the turns as I unwound it, I cannot state its ampere-turn value. I suspect that it is a good bit less than 1200 when operated on 12V.

 

Dave,

 

If you can be bothered, just count the number of turns on the top layer and multiply by the 14 layers I wind on them.